10xDev

Relax, go outside, touch grass. The sun will rise again.

Severely misleading statement.

People have already forgotten about Gaza/W bank and there is a ground invasion of Lebanon happening that no one is talking about. They are absolutely winning if their goal is regional dominance.

No just chronic sleep deprivation. <6 hours sleep a day catches up to you and the effects can be very severe.

Relax, it is one demo. It probably can't handle the millions of edge cases that exist in real life.

It is somewhat integrated to Instagram with billions of active users. Not saying a whole lot.

>Europe won't be able to catch that. If we cut the chip supply right as things take off, China might not either.

https://www.asml.com

Because it hasn't happened yet.

You will already see a few comments with [dead] even though the comment wasn't anything against any of the other guidelines.

They are attempting MAD but without nukes. The problem is that while the US seems to have obliterated Iran, they have no endgame plan on actually accomplishing a regime change. They thought about arming Kurdish Iranians but wen't against the idea, now we seem to be in a perpetual state of "it is over / it is not over".

>capitalisation is just unimportant. Signing off emails with "best wishes" is not a good use of anyone's 500 milliseconds

Yet I'm on Twitter reading "Prison for attempted murderer enablers like this clown" by the world's richest man who is tweeting all day. My guess is that it has just become a way of status signalling more than anything else.

They are going to be missing out on top content https://www.instagram.com/popular/ai-generated-video-shorts/

Those sound more like emergent properties then something you can engineer.

You will get laid off but you won't if you create value and you will create value if you don't care about money and it will go recursive but it won't go recursive because there are limits. It won't change anything but it will change everything.

No one knows what he is actually saying (see comments) but at least he managed to compress the entire discourse on AI impact into a blog post.

A lot of juniors only graduated using these tools. Good luck taking it away from them.

With AI it makes sense to have leaner teams. Being able to go faster requires greater responsibility.

Stochastic and unpredictability aren't exactly the same. I would claim current LLMs are generally predictable even if it is not as predictable as a deterministic program.

There is still a huge gap between a model continuously updating itself and weekly patches by a specialist team. The former would make things unpredictable.

The fact that models aren't continually updating seems more like a feature. I want to know the model is exactly the same as it was the last time I used it. Any new information it needs can be stored in its context window or stored in a file to read the next it needs to access it.

Whether it is text or an image, it is just bits for a computer. A token can represent anything.

Already loads of posts on this. It is mostly theatre.

They have researchers working for insane salaries just so they don't go to another frontier lab to share their ideas. If you think it is just "mindless bruteforce" you don't understand anything. The idea is that the most effective methods are ones that scale but those ideas are also then limited by the compute available.

Seems like no matter what people will continue to make reductive statements like this.

People are finding it hard to grasp emergent properties can appear at very large scales and dimensions.

CoT already moved things past the "it is just token prediction" phase. We have models that can perform search over a very large state space across domains with good precision and refine its own search leading to a decent level of fluid intelligence, hence why ARC AGI 1/2 is essentially solved. We also don't know the exact details of what is happening at frontier labs seen as they don't publish everything anymore.

With Gemini Pro on Antigravity you get a quota reset every 5 hours and access to Claude Opus 4.6. That's what I use at home and don't need anything else.

It seems to already be helping with open problems: https://www-cs-faculty.stanford.edu/~knuth/papers/claude-cyc...

That's because in the 1% of cases it fails it could result in someone dying. In fields where there isn't the same level of risk or regulation involved it shouldn't be as resistant to change.

The next revolution is coming and it is well needed. Society is becoming older, more tired and we need new fresh ideas to bring a lot of fields back to life. I hope it comes soon.

The goal posts are becoming more narrow and these posts are becoming more frequent. It is almost like a therapy session for those facing an existential crisis while they continue to train the very thing that will replace them by giving it more training data to do their work.